融合点测量信息的光学层析方法

Algebraic Reconstruction Technique Fused with Pointwise Measurement

光学计算机层析术的困难在于不能得到足够的信息以完全确定探测目标。理论上 ,将先验知识融入层析术可以在一定程度上降低病态程度 ,而实际中场的一组孤立点的测量值是一种更为直接的可得信息。利用 ART3算法对于含噪声、甚至部分缺失的投影数据具有很好的鲁棒性 ,并且能友好地融合来自其他测量系统的信息的特点 ,提出一种利用点测量信息的自适应方法 ,避免了对重建过程引入原本不存在的高频信息 ,以及在迭代过程中对预期的低频分量的过抑制 ,保证系统的稳定。

These isolated point measurements deserve pertain roles in the OCT reconstruc- tion.The ART3algorithm is a robust method dealing with practically noisy or even partly missing projections,and is friendly to integrate the information from various sources.Most of the suggestions on fusing information inevitably introduce high frequencies,which never exist in the field,into the interation. And these high frequencies over- suppress the expected low frequency components and cause the reconstraction non- stable.An adaptive way is de- veloped to take a good use of pointwise priori.By analyzing the simulated experiments,some deserved attention to the application of priori in OCT is strongly suggested.